relationship between PI_HAT and --king-cutoff coefficient?

Question

[anamaria]

Hello,

as I see from plink documentation to keep only distantly related samples PI_HAT=P(IBD=2)+0.5*P(IBD=1)=0.125 according to this table 1 https://www.sciencedirect.com/science/article/pii/S0002929714004273

but I am unable to find a table which would compare plink --king-cutoff coefficients with PI_HAT explicitly

My guess from what I read is that --king-cutoff 0.0884 would correspond to PI_HAT = 0.125

Can someone please give me some input on this?

Answer 1

[Sam]

The king cutoff is based on a software called KING. You can find more information here: https://people.virginia.edu/~wc9c/KING/manual.html

Based on the document:

an estimated kinship coefficient range >0.354, [0.177, 0.354], [0.0884, 0.177] and [0.0442, 0.0884] corresponds to duplicate/MZ twin, 1st-degree, 2nd-degree, and 3rd-degree relationships respectively

[anamaria]

Thanks for sharing that. One thing is unclear to me is it necessary to run: plink --bfile output8 --genome --grm-cutoff 0.1 --make-bed --out output9

after I did run this bellow? plink2 --bfile output7 --king-cutoff 0.0884 plink2 --bfile output7 --remove plink2.king.cutoff.out.id --make-bed --out output8

[Sam]

So you ran plink2 --bfile output7 --king-cutoff 0.0884 plink2 --bfile output7 --remove plink2.king.cutoff.out.id --make-bed --out output8?

where do you get the plink2.king.cutoff.out.id file?

If you have performed king filtering, I don't think you need to do the grm filtering.

Side note: You might want to prune your data before you perform king or grm fitlering

Answer 2

[lud_0911]

As far as I know, --king-cutoff 0.08 would correspond to PI_HAT = 0.2 because both filter for second-degree relationship. But I tried both in my dataset and diferent individuals were removed. The last removed more than the former. Does anyone know the reason?

[chrchang523]

If you need to ask, then you should ignore PI_HAT. A major advantage of the KING method is that it is much harder to misuse.